Project 1. Molecular mechanisms of multidrug resistance in the pathogenic yeast, Candida glabrata. Inherent resistance in C. glabrata to azole antifungal drugs and the rapid increase of this resistance during azole therapy has seriously limited the efficacy of this important class of drugs . C. glabrata is second only to Candida albicans as a cause of Candida bloodstream infections and is particularly commonly isolated from the blood of patients already receiving an azole. The major mechanism of azole resistance in this species appears to be efflux mediated by multidrug resistance (MDR) genes. Mutations in the azole target, C14 demethylase, appear to be less important. Research in the Clinical Mycology Section has centered around the drug efflux genes in C. glabrata and their transcriptional regulation. CgAP1, a transcriptional regulatorThe Saccharomyces cereviseae gene, YAP1, codes for one of 8 basic leucine zipper (bZIP) transcriptional regulatory proteins in yeast. The function of YAP1p appears to be protecting the cell from stress, including oxidation and toxic chemicals. Interrupting either YAP1 or the related Candida albicans gene, CAP1, increases susceptibility to a broad array of chemicals, including cadmium (Cd), 4 nitroquinolone-N-oxide (NQO), 1,10-phenanthroline (Ph), hydrogen peroxide (H2O2) and cycloheximide (Cx). Target genes of YAP1 include both drug efflux pumps and genes controlling intracellular glutathione stores. Overexpressing CAP1 in a fluconazole susceptible (pdr5) strain of S. cerevisiae has been reported to increase fluconazole resistance, an effect mediated by the major facilitator efflux pump, FLR1p. In order to understand the role that homologues of YAP1 and FLR1 might play in the acquisition of azole resistance in C. glabrata, both genes were cloned and sequenced using sequence homology with the S. cerevisiae genes. Function of CgAP1 was explored by interrupting the gene in two C. glabrata hosts, gene replacement in one of the interrupted C. glabrata strains, complementing the Ayap1 mutation in a S. cerevisiae host, and overexpressing CgAP1 in S. cerevisiae. Results were as follows.CgAP1 was found to complement YAP1 by restoring resistance to H2O2, NQO and Cd in a Ayap1 strain of S. cerevisiae. As reported last year, overexpressing CgAP1 under control of its own promoter in a multicopy plasmid increased resistance to NQO, Ph, diamide, and Cx. In subsequent work, no single target drug efflux gene leading to this increased resistance could be found, using strains with single deletions of PDR5, YCF1, FLR1 and ATR1. Overexpressing CgAP1 in a fluconazole susceptible (pdr5) S. cerevisiae host increased fluconazole resistance 32 fold. Interruption of CgAP1 in two C. glabrata strains using different constructs (producing strains Acgap1::URA3 and Acgap1::LEU2,ura3) caused increased susceptibility to H2O2, NQO and Cd but not to fluconazole. In the Acgap1::LEU2, ura3 strain, gene replacement with CgAP1 in a low copy plasmid restored resistance against H2O2, NQO and Cd to that of the parental strain. Interruption of CgFLR1 had no observable phenotype. As in S. cerevisiae, benomyl increased transcription of FLR1, an effect that was blocked in a Acgap1 strain, meaning that CgFLR1 is one of the targets of CgAP1. When five pairs of fluconazole susceptible or resistant C. glabrata isolates were compared, the isolates being otherwise congenic, no difference in FLR1 transcript abundance was found. We reported last year that CgYAP1 transcript abundance also did not differ between these paired clinical isolates. In sum, CgAP1 and CgFLR1 were found to have similar but not identical function in C. glabrata to that of YAP1 and FLRI in S. cerevisiae. Neither gene appeared to play a major role in fluconazole resistance among the clinical isolates we studied.PDH1, an ATP-binding cassette (ABC) transporterCloning and sequencing of this gene was described in an earlier annual report. In subsequent work, the PDH1 gene, which codes for an ABC transporter in Candida glabrata, was found to provide functional complementation of a Saccharomyces cerevisiae null pdr5 mutant. Disruption of PDH1 altered drug susceptibility in a pattern similar to that of pdr5 mutants. Disruption of PDH1 in a Acgdri::URA3 background increased susceptibility to rhodamine 6G, cycloheximide and chloramphenicol, and also increased rhodamine 6G uptake. Expression of PDH1 in a C. glabrata Acgcdr1, Apdh1 mutant reversed the phenotype of both mutations, indicating that the phenotypes conferred by these two genes overlap. Abundance of PDH1 and CgCDR1 transcripts in C. glabrata was increased by rhodamine 6G, cycloheximide and oligomycin. Expression of PDH1 in S. cerevisiae complemented the pdr5 mutation by reversing susceptibility to rhodamine 6G, chloramphenicol and cycloheximide, as well as by decreasing rhodamine 6G intracellular concentrations. Phenotypes of S. cerevisiae yor1 and snq1 mutations were not restored.These findings show that the C. glabrata gene PDH1, like CgCDRI, is a homologue of the S. cerevisiae gene, PDR5.Deletion of the CgCDR1 in C. glabrata has been reported to increase fluconazole susceptibility. Frequency of fluconazole resistant mutants has been reported to be less in a Acgcdr1, Apdh1 than in a Acgcdr1 strain. Whether the in vitro findings of increased susceptibility or fewer resistant mutants has significance during an infectious process has not been studied. We determined that Acgcdr1 and Acgcdr1, Apdh1 strains had virulence equivalent to the host strain, NCCLS84 in a gp91phox(-/-) knockout mouse. As reported last year, C. glabrata causes progressive, lethal infection in this, mouse strain but not in the C57Bl/6 parental strain. Knockout mice were injected intravenously with C. glabrata and treated with fluconazole 160 mg/kg per day intraperitoneally for six days. Splenic fungal burden was measured two days later. Fluconazole had no effect on fungal burden of mice infected with NCCLS84 but caused a significant reduction of splenic fungal burden in mice infected with the more fluconazole susceptible Acgcdr1 strain. No additional reduction in splenic burden was found in mice infected with the double deletant, Acgcdr1, Apdh1strain. Deletion of the ABC drug transporter CgCDR1 was important for conferring fluconazole susceptibility in this mouse model but additional deletion of PDH1 conferred no additional responsiveness to fluconazole treatment. Project 2. Multicenter, randomized comparison of voriconazole with amphotericin B for the primary treatment of invasive aspergillosis.The Clinical Mycology Section head was one of the leaders in a multinational study of invasive aspergillosis. This investigator assisted in the following efforts: writing the protocol, analyzing the data during the study as a member of the data review committee, analyzing the final data and writing the manuscript as a member of the writing committee. The results were as follows. Patients were randomized to receive intravenous voriconazole (6 mg per kilogram for two doses on day 1 then 4 mg per kilogram twice daily) followed by 200 mg twice daily orally or amphotericin B deoxycholate (1 mg per kilogram per day). Other licensed antifungal treatments were allowed after failure or intolerance to initial randomized therapy. Successful outcomes were complete or partial responses. Endpoints were response at week 12 and at end of initial randomized therapy and survival through week 12. Results were as follows: 144 voriconazole and 133 amphotericin B patients had definite or probable aspergillosis and received at least 1 dose of treatment. Underlying conditions were allogeneic hematopoietic transplantation, acute leukemia or other hematological diseases, solid organ transplantation, AIDS and corticosteroid therapy. At week 12, successful outcome occurred in 52.8 percent (20.8 percent complete responses and 31.9 percent partial responses) voriconazole and 31.6 percent (16.5 percent complete responses and 15 percent partial responses) amphotericin B treated patients (95 percent confidence interval for the difference in successful outcome: 10.4 to 32.9 percent). Survival was 70.8 percent in voriconazole patients and 57.9 percent in amphotericin B patients (Hazard Ratio 0.59; 95 percent confidence interval 0.40 to 0.88). Voriconazole treated patients experienced fewer drug-related adverse events (p=0.025). The authors concluded that patients begun on voriconazole for invasive aspergillosis showed a better response, improved survival and suffered fewer side effects than those begun on amphotericin B.